Weathering change: Longitudinal patterns of residential mobility following extreme weather events in Australia

• The odds of moving increase up to three years after home damage caused by a flood, bushfire or cyclone. • We identify four types of climate-induced relocation: immediate, temporary, delayed and repeat. • Extreme weather events bring forward some moves. Fewer moves than expected are recorded the year after. • Distributed lag models provide nuanced insights into the spatiotemporal patterns of climate-induced population movement. There is growing evidence of disaster-induced mobility. However, most research focuses on immediate displacement, limiting our ability to understand the long-term consequences of extreme weather events on residential mobility. To address this gap, this study establishes the relationships between home damage caused by bushfires, floods, and cyclones and ensuing residential mobility in the following five years. Using nationally representative individual-level data from the Household, Income and Labour Dynamics in Australia (HILDA) Survey from 2009 to 2023, we use finite distributed lag logistic models to determine the level, timing, and direction of weather-induced mobility, distinguishing between onward and return movement and distance moved. While most moves occur in the immediate aftermath of an extreme weather event, results indicate that weather-induced home damage increases the odds of moving for up to three years. Four distinct mobility patterns emerge: (1) immediate relocation , which brings some moves forward, (2) delayed displacement up to three years later, particularly over short distances, (3) short-term displacement followed by return, and (4) repeat onward migration . Younger adults and renters are more likely to be involved in all mobility types. Tertiary-educated individuals are less likely to return, but more likely to move repeatedly, while separated individuals are more likely to be delayed movers. The diversity of mobility responses underscores the conceptual and methodological complexities of estimating weather-induced population movement and the need for nuance and caution, while showing the benefits of distributed lag models to capture time dynamics. Our findings also highlight the need for targeted support mechanisms that account for different mobility trajectories and varying constraints faced by sub-population groups in the aftermath of extreme weather events.